Field of the Invention
The invention relates generally to the field of manufacturing integrated semiconductor circuits, such as VLSI and ULSI circuits, using photolithographic methods. In particular, the invention relates to the increase of the resolution capacity of conventional photolithography through the use of alternating phase masks.
In the manufacturing of integrated semiconductor circuits, the mask structures allocated to the circuit elements are optically imaged on light-sensitive layers on the wafer in a conventional manner. Due to the diffraction effect, the resolution capacity of such an imaging system is limited, and mask structures having dimensions below the reciprocal value of this resolution capacity, called the critical structures, are imaged in blurred or unfocused fashion. This leads to undesired strong correlations of the circuit elements, and thus to an adverse effect on the circuit functionality.
These difficulties can be overcome by exploiting the destructive interference effect of two closely adjacent and coherent light beams that are phase-displaced by 180xc2x0, and converting the conventional masks concerned into alternating phase masks, in which each critical structure is provided with two phase shifters in order to produce the required phase displacement.
The various types of phase masks are for example described in the book xe2x80x9cTechnologie hochintegrieter Schaltungen,xe2x80x9d by Widmann, Mader, and Friedrich, 2nd ed., Springer-Verlag, pp. 135ff. A detailed overview of phase mask technology is contained in the publications xe2x80x9cImproving Resolution in Photolithography with a Phase-Shifting Mask,xe2x80x9d by M. D. Levenson et al., in IEEE Trans. Electron. Devices 29 (1982), 1828ff., and xe2x80x9cWavefront Engineering for Photolithography,xe2x80x9d by M. D. Levenson 993, in Physics Today, Jul. 1993, p. 28ff.
The use of so-called strong phase masks, including both the already-mentioned alternating phase masks and also phase masks without chromium, requires that in each affected level, the transparent phase-shifting structures are assigned to one of two phases, having a phase difference xcex94"PHgr"=180xc2x0. Here the following two cases must be distinguished. In what is known as a dark-field phase mask, transparent structures correspond to the circuit elements (e.g., printed conductors), and phases can be assigned to them, while opaque mask fields are formed by regions covered with chromium. In contrast, in a so-called bright-field phase mask the opaque regions, covered with chromium, of the phase masks represent the circuit elements, and the regions situated between them are transparent. In the latter case, suitable regions in the vicinity of the opaque chromium regions must be designated as phase-shifting elements. The creation of the phase-shifting elements takes place according to particular design rules known from the prior art, and is described for example in U.S. Pat. No. 5,537,648.
However, in view of the complexity of modern circuits, and the requirement of two elements that are phase-displaced by 180xc2x0 to one another at each critical structure, phase conflicts are conceivable. A phase conflict is present precisely when the phase shifters on both sides of a critical structure are erroneously assigned the same phase, or when, due to the interaction of the phase-shifting elements, the destructive interference effect occurs at an undesired location on the already-mentioned light-sensitive layer. The phase assignment for the different phase-shifting elements thus represents a mathematical-combinatorial problem that does not have a general solution. Because in principle the phase assignment can lead to different results, and different phase assignments can occur for one and the same cell of a hierarchical layout, the phase assignment in an automated program must finally be carried out at the finished circuit layout. An automated checking routine is therefore required that examines a circuit layout in order to see whether a phase assignment is possible at all. This check should be complete, and should localize the problem point as well as possible, i.e., should determine its actual point of origin. The latter is not self-evident, because the location at which it is discovered that the combinatorial problem does not xe2x80x9cwork outxe2x80x9d may be located far from the actual point of origin.
After phase conflicts have been determined in an automated routine, they can be resolved in two fundamentally different ways. First, the circuit design can be modified slightly at the locations of the localized phase conflicts, for example by displacing printed conductor structures, so that the phase conflicts are removed. On the basis of this modified circuit design, a successful phase assignment can then be carried out in order to create a phase mask. However, this method is avoided whenever possible, because it is not in accord with the miniaturization process, governed by Moore""s Law, in microelectronics.
Second, the circuit design can remain unmodified, and instead phase conflicts can be solved by assigning two different phases to individual phase-shifting elements. However, this has the consequence that during exposure a dark region occurs at the boundary line between the two different phase regions, which would lead to an interruption. For this reason, in this case an additional exposure step must be carried out using what is called a trim mask, by means of which the occurrent dark boundary regions are separately exposed at a later time. However, in such a subsequent exposure using a trim mask, there is then in turn the danger that, if the dimensions of the trim mask structure are below the reciprocal value of this resolution capacity of the exposure system, these photolithographically critical structures will be imaged in blurred or unfocused fashion, which can lead to an adverse effect on the functionality of the integrated semiconductor circuits.
In the prior art, various methods are known for testing a layout for phase conflicts. In the publication xe2x80x9cHeuristic Method for Phase-Conflict Minimization in Automatic Phase-Shift Mask Design,xe2x80x9d by A. Moniwa et al., in Jpn. J. Appl. Phys., vol. 34 (1995), pp. 6584-89, a graph-theoretical approach is described in which a set of phase-shifted elements is assumed, and from this set a planar undirected graph is formed, taking into account the technological requirements. In addition, U.S. Pat. No. 5,923,566 describes a computer-implemented routine that verifies whether an existing circuit design can be imaged onto a phase mask, or whether localized phase conflicts are present. However, neither of the methods described above works optimally in the acquisition of phase conflicts. First of all, both these methods prove to be inefficient, because in them for example certain phase conflicts are indicated twice. Second, they prove to be inadequate, because with them certain other phase conflicts cannot be acquired. In addition, in the cited references such phase conflicts are not resolved, and moreover the problem is also not addressed of an insufficient subsequent exposure of dark boundary regions with the known trim mask structures.
U.S. Pat. No. 5,523,198 describes a method in which boundary sections that were not sufficiently exposed in a first exposure step using a phase mask are subsequently exposed in a second step using a pure light-dark mask. In addition, a method is described in which, in two successive exposure steps, phase masks and a trim mask are used. Such trim masks are additionally known from the international PCT publication WO 2001/06320 A1 and from Japanese patent application JP 2000-338637 A.
It is accordingly an object of the invention to provide an improved method for determining possible phase conflicts on alternating phase masks, which overcomes the above-mentioned disadvantages of the heretofore-known devices and methods of this general type and which allows for the automatic removal of these phase conflicts. It is a further object to provide a trim mask with which an exposure is ensured even of photolithographically critical dark structures.
With the foregoing and other objects in view there is provided, in accordance with the invention, a method for exposing a photoresist layer on a substrate wafer, which comprises the following method steps:
in a first exposure step, exposing the photoresist layer in first sections with a first exposure radiation and in second sections with a second exposure radiation, the first exposure radiation being phase-displaced by 180xc2x0 relative to the second exposure radiation, and the first and second sections adjoining one another in boundary regions that are not sufficiently exposed;
determining first and second boundary regions respectively having less than a given predetermined least distance from one another; and
in a second exposure step, reexposing the photoresist layer in the first boundary region with a third exposure radiation, and in the second boundary region with a fourth exposure radiation, the third exposure radiation being phase-displaced by 180xc2x0 relative to the fourth exposure radiation.
In other words, the objects of the invention are achieved with the inventive method for exposing a photoresist layer on a substrate wafer in which, in a first exposure step in which the photoresist layer is exposed in a first section with a first radiation and in a second section with a second exposure radiation, whereby there is a phase displacement of 180xc2x0 between the first exposure radiation and the second exposure radiation. The first and second sections thereby adjoin one another in boundary regions that are not sufficiently exposed in the first exposure step. After the determination of first and second boundary regions each having less than a particular predetermined least distance, i.e., photolithographically critical distance, from one another, in a second exposure step a subsequent exposure of the photoresist layer is then carried out in the first boundary region with a third exposure radiation, and is carried out in the second boundary region with a fourth exposure radiation, there being a phase displacement of 180xc2x0 between the third exposure radiation and the fourth exposure radiation. The remaining boundary regions can here be exposed with a further exposure radiation having arbitrary phase.
The advantage of the invention is that after the use of a phase mask even after the second exposure step, there remain no regions in the photoresist layer that are exposed in defective fashion due to diffraction effects. Defectively exposed regions in the second exposure step occur when two boundary regions that are to be subsequently exposed have a distance from one another that is so small in relation to the wavelength of the exposure radiation that undesirable diffraction effects occur during imaging.
Such diffraction effects occur in a conventional trim mask having structural dimensions that are smaller than the reciprocal value of the resolution capacity in the imaging, because in this case these photolithographically critical structures are imaged in blurred or unfocused fashion. This can lead to undesired strong correlations of the circuit elements, and thus to an adverse effect on the circuit functionality. In order to avoid this, according to the invention a phase shifting of the exposure radiation is provided also in the second exposure step, so that with the aid of the destructive interference effect, which is already exploited for the imaging of structures in the first exposure step, a defective exposure of the boundary regions that are to be subsequently exposed can be avoided. With the trim mask, fashioned as an alternating phase mask, through the destructive interference effect of the third and fourth exposure radiation with opposite phases all photolithographically critical structures at the boundary regions that are to be subsequently exposed are correctly imaged.
Advantageously, the exposure steps are carried out with the aid of a phase mask, the phase mask having regions that are transparent to light, which assigns a particular phase shift to the light passing through.
According to a preferred specific embodiment, if the inventive method is to be applied in a dark-field phase mask technology, the mask for the first and/or second exposure step is determined as follows:
In a first method step, critical regions are determined in which two adjacent, transparent regions provided for the phase mask have less than a particular predetermined least distance from one another.
In a second method step, overlapping regions are determined between rectilinear segments of the obtained critical regions and end regions of rectilinear segments, ending in the middle of transparent regions, of the critical regions, and degenerate critical regions are generated. The latter are obtained by subtracting overlapping regions from the critical regions.
In a third method step, coherent regions (xe2x80x9ccountriesxe2x80x9d) located outside the transparent regions and outside the critical regions are then determined, as are large outer borders of the countries and of the overlapping regions and end regions obtained in the previous method step.
Then, in a fourth method step, from each of the determined outer borders the number of segments of contact with the degenerate critical regions is determined, and, if the number is odd, a phase conflict is determined.
Finally, in a fifth method step, the phase conflicts are resolved by determining coherent layout regions and the region borders thereof, coherent layout regions being defined by the transparent regions and the critical regions located between them minus overlapping regions representing the end regions and the phase conflicts, and at least one connection path being produced between an outer border representing a phase conflict and either a nearest additional outer region border or an odd number of not-yet-connected large outer borders, which likewise represent phase conflicts; then the quantity of produced connection paths is reduced to those in which each phase conflict is contained exactly once, and then those regions (covering regions) of the connection paths are marked that are located over the transparent regions, and finally, for the phase mask the covering regions are fashioned as boundaries between two different regions of the phase mask to be manufactured, whose phase displacements have a phase difference of xcex94"PHgr"=180xc2x0.
According to a preferred specific embodiment, if the inventive method is to be applied in a bright-field phase mask technology, the mask for the first and/or for the second exposure step is determined as follows:
In a first method step, phase-shifting regions are respectively determined on both sides of opaque critical regions provided for the phase mask. Critical regions are defined in that they have less than a predetermined structural width.
Next, in a second method step overlapping regions are generated between rectilinear segments of the critical regions and end regions of rectilinear segments, ending in the middle of phase-shifting regions or regions of interaction between phase-shifting regions, of the critical regions, and degenerate critical regions are generated. The latter are obtained by subtracting overlapping regions from the critical regions.
In a third method step, coherent regions (countries) located outside the phase-shifting regions, regions of interaction, and critical regions, are determined, as are large outer borders of these countries, and of the overlapping regions and end regions obtained in the previous method step.
Finally, in a fourth method step, from each of the determined outer borders the number of segments of contact with the degenerate critical regions is determined, and if the number is odd, a phase conflict is determined.
In a fifth method step, the phase conflicts are then resolved in that coherent layout regions and the borders thereof are determined, coherent layout regions being defined through the phase-shifting regions, the critical regions, and the regions of interaction minus regions of overlap representing the end regions and the phase conflicts, and at least one connection path being produced between a large outer border representing a phase conflict and either a nearest additional external region border or an odd number of not-yet-connected large outer borders, which likewise represent phase conflicts; then the quantity of produced connection paths is reduced to those in which each phase conflict is contained exactly once, and then those regions (covering regions) of the connection paths are marked that are located over the transparent phase-shifting regions, and, finally, for the phase mask the covering regions are fashioned as region boundaries between two different regions of the phase mask to be manufactured, whose phase displacements have a phase difference of xcex94"PHgr"=180xc2x0.
According to the two inventive developments presented above, in a first step thereof a formalism is thus provided by means of which the direct convertibility of integrated semiconductor circuits into alternating phase masks B both dark-field and bright-field phase masks B is tested. This takes place through an explicit localization of the phase conflicts occurring in the corresponding layout, with the application only of the technical demands placed on the design. The set of phase conflicts determined with the aid of this formalism is complete and minimal, i.e., all existing phase conflicts are always determined, and existing phase conflicts are not indicated more than once.
Subsequently, the phase conflicts are removed through the described automated method. In the removal of the phase conflicts, both in dark-field and also in bright-field technology the connection paths are formed practically and rapidly in such a way that at first pairs of opposite edges of a respective large outer border and of an adjacent outer region border are determined, and subsequently at least one connection path is produced between the edges of each pair.
Since, however, the connection paths must later lead to phase jumps, i.e., sudden phase changes in the phase-shifting regions, in the group-by-group solution of the associated neighborhood problem only those edges are to be taken into account that are to be allocated to the large outer contour of the respective layout group and to the polygons representing phase conflicts, from which the contact segments are to be subtracted.
However, the connection paths do not necessarily run between edges of pairs of large outer borders representing the phase conflict and the layout group borders. Moreover, a connection path between two phase conflicts, or between a phase conflict and the outer contour of a layout group, need not necessarily be even; rather, it can run through countries having an even number of contact segments (i.e., no phase conflicts), and can assume a very complicated form. Its essential feature is that it connects a phase conflict with the additional external outer contour of its layout group, or connects two phase conflicts with one another.
After the production of the connection paths, these are reduced to a quantity of connection paths in which each phase conflict occurs only once. For the inventive removal of the phase conflicts, it is thus sufficient respectively to retain only one connection path between two phase conflicts, or between a phase conflict and the outer contour of its layout group, after the reduction.
Subsequently, those regions of the connection paths that are located over the phase-shifting regions are taken into account or are marked in some way. Then, for the preparation for the manufacture of the first phase mask, the covering regions are formed as boundaries between two different regions of the phase mask to be manufactured. For this purpose, for example the covering regions can be subtracted from the phase shifters, and the regions of the first phase mask to be manufactured can be represented in a suitable matter, e.g., as two different colors of a two-color phase mask. The present invention describes a double exposure technique for the automatic handling of arbitrary layouts in phase mask technology. The mask for the second exposure process is thereby likewise a phase mask. This enables the handling of propagating phase conflicts. The second phase mask is automatically generated from the phase-shifting elements, the structures, and the connection paths between the borders, calculated using the above-explained formalism, for the automatic removal of phase conflicts.
In addition, according to the invention a trim mask is provided for use in the above-named method, with which a photoresist layer is exposed in a second exposure step after an exposure with a phase mask. The trim mask has a first region transparent to light and a second region transparent to light, in order subsequently to expose the regions in the photoresist layer that were artificially unexposed through the use of the phase mask in the first exposure step. The first light-transparent region and the second light-transparent region are formed such that the phase displacement of 180xc2x0 is obtained between the third exposure radiation, passing through the first transparent region, and the fourth exposure radiation, passing through the second transparent region. The isolated trim openings can be exposed with an additional exposure radiation having arbitrary phase. This trim mask has the advantage that it is suitable for the subsequent exposure of two regions on the photoresist layer that were artificially not exposed in a first exposure step. In particular in the propagation of phase conflicts, it happens that the artificially unexposed regions are located on both sides of a critical structure, and thus have a spacing that is too small for a precise imaging using conventional masking technology. Given a conventional trim mask, one would thus obtain a defective exposure between the subsequently exposed regions. In contrast, the inventive trim mask has the advantage that, due to the destructive interference effect in the exposure of the artificially exposed regions in the second exposure step, it enables an exact imaging of the critical structure. The correction of the interferences through the boundary regions between the phase-shifting regions of the phase mask of the first exposure step thus does not have the result that the structure to be imaged is represented in unfocused fashion due to diffraction effects, or is destroyed by the second exposure step.
Advantageously, only those regions are provided with a 180xc2x0 phase displacement that relate to photolithographically critical structures, while all uncritical structures are exposed with the aid of a transparent region, without a fixedly predetermined phase displacement.
In the context of bright-field phase mask technology, a trim mask is thereby preferably used that consists of trim structures, a dark region, and an associated bright background, the trim structures being colored in such a way that those trim structures located at critical structures overlapped by connection paths receive opposed phases for the exposure radiation, whereby a phase displacement of 90xc2x0 to the bright background is maintained.
For application in the context of dark-field phase mask technology, the trim mask preferably consists of trim structures, a bright region, and an associated dark background, the trim structures being colored in such a way that those trim structures located at critical structures overlapped by connection paths receive opposed phases for the exposure radiation.
The trim structures are thereby determined, both in bright-field and also in dark-field phase mask technology, by parallel expansion along the phase shifters of the above-determined covering regions.
The two exposure masks in bright-field or dark-field phase mask technology can be divided into the following regions: critical structures (KR); uncritical structures (UKR); undecomposed phase shifters (PO); connection paths (VW) from the automatic removal of phase conflicts; segments (VWPO) of the connection paths over the undecomposed phase shifters (PO); segments (VWKR) of the connection paths over the critical structures (KR); artificially decomposed phase shifters (PZ) that correspond to the undecomposed phase shifters (PO) minus segments (VWPO) of the connection paths over the undecomposed phase shifters (PO); trim structures (TS) that are determined by extending the segments (VWPO) of the connection paths over the undecomposed phase shifters (PO) parallel to the undecomposed phase shifters (PO) to the left and to the right by a predetermined length, called the trim width; and regions (TBG) composed from undecomposed phase shifters (PO), critical structures (KR), uncritical structures (UKR) minus the trim structures (TS).
In addition, the multiply coherent region outside the phase shifters and additional structures is designated the background, whereby it is not distinguished whether these structures are photolithographically critical or uncritical. In bright-field technology, this background is bright, and in dark-field technology it is dark.
The two exposure masks, i.e., the phase mask (M1) for the first exposure step and the trim mask (M2), are automatically determined, according to the method presented above, as follows:
In dark-field phase mask technology, phase mask (M1) is made up of a two-colored set of artificially decomposed phase shifters (PZ) and the associated dark background. Trim mask (M2) for the subsequent exposure of artificially produced dark regions on the photoresist consists of a two-colored set of trim structures (TS), regions transparent to light in the uncritical structures (UKR.), and the associated dark background.
In bright-field phase mask technology, phase mask (M1) is made up of a two-colored set of artificially decomposed phase shifters (PZ) and the associated bright background. Trim mask (M2), for the subsequent exposure of artificially produced dark regions on the photoresist, consists of a two-colored set of trim structures (TS), dark composed regions (TBG) representing the undecomposed phase shifters (PO), the critical structures (KR), and the uncritical structures (UKR) minus the trim structures (TS), and the associated bright background, having a 90xc2x0 phase difference from the trim structures (TS)
Trim structures are thereby always colored such that two trim structures on a segment (VWKR) of the connection path over the critical structures (KR) receive opposed phases. All other trim structures can be exposed with an arbitrary phase. This holds also for regions transparent to light on the uncritical structures (UKR). However, a matching to one of the phases in the colored trim structure region is preferably carried out. If this trim structure region has for example a 90xc2x0/270xc2x0 coloring, then 90xc2x0 or 270xc2x0 phase is selected for the remaining trim structures and regions transparent to light at uncritical structures. In addition, in trim structure coloration in bright-field phase mask technology, a 90xc2x0 phase difference to the bright background is maintained, because given a 90xc2x0 phase difference the light intensity lies above the threshold of sensitivity of the photoresist, so that unexposed regions do not occur on the edge of composed regions (TBG). If, for example, 90xc2x0/270xc2x0 coloration is used for the trim structures, a 180xc2x0 phase is assigned to the bright background.
Other features which are considered as characteristic for the invention are set forth in the appended claims.
Although the invention is illustrated and described herein as embodied in a method for determining and removing phase conflicts on alternating phase masks, and mask configuration for use with such a method, it is nevertheless not intended to be limited to the details shown, since various modifications and structural changes may be made therein without departing from the spirit of the invention and within the scope and range of equivalents of the claims.
The construction and method of operation of the invention, however, together with additional objects and advantages thereof will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings.